Coherence in Punjabi: Discourse Oriented Generalizations


Author: Harjit Singh (Center for Linguistics, School of Language Literature & Culture Studies, Jawaharlal Nehru University, India)
Speaker: Harjit Singh
Topic: Anthropological Linguistics
CALA 2020 General Session


Abstract

This paper studies relational mechanism between segments of a text in terms of analyzing its soundness or organization. Basically formal as well as cognitive experiments proved with various kinds of consistency or logical sequence of a text. And the eventual goal of such experiments to check the all-possible but legitimate interpretations of verbal or sometimes non-verbal communication forms in a language. With such inquiries, we also take this initiative in Punjabi and try to incorporate them with some findings here. ‘Coherence’ generally means to indicate the meaningful or understandable interpretation of any given oral as well as written dialogue or consider it as a parameter to know discourse processing from production to comprehension. In fact it is the blend of textual and mental activity of human, which is analyzed with various perspectives. In short, at written level, it counts with utterances, dialogues, or more clearly segments of a discourse structure. Grosz and Sidner (1986) emphasized on ‘intentional mode of a speaker’ to understand coherence relations in a dialogue. While Hobbs (1980, 82), Marcu (2000) preferred ‘informational mechanism’ for such relations and also tried to use this knowledge in natural language processing.

We divide this work into five sections. The concept of ‘coherence’ and its basic taxonomy of discussed in the first section with special reference to works of (Hobbs 1985, Webber et al. 2003) in ‘cause-effect relations’ and again its sub-typology as and then we further focused on ‘volitional and non-volitional typology’ of Marcu (2000) also. While in the second section, we give some overview about Punjabi, which is Modern-Indo-Aryan tonal language and it has also followed (SOV) word order with head final construction. In the third section, we take two types of analysis in Punjabi, particularly focus on Punjabi ILCI corpora data files. First we check those ‘contentful conjunctions’ from the informational mechanism point of view. Secondly we take Hobbs (1985) suggested labeling of coherence relations in Punjabi. In the fourth section, we discuss the applicability of used above two kinds of informational methods to finalize the status of Punjabi ILCI corpora data files in terms of coherent content. In a similar way, we evaluate this work with related issues and proposed hypothesis in the last section.

Keywords: Coherence, intentional, informational, discourse segments etc.